RobustVectorAlignment

Overview

This repository provides the implementation and resources for Spherical Point Pattern Registration algorithms developed as part of Experiment 1. These algorithms aim to robustly align vector fields on the unit sphere.

Directory Structure

• vector_alignment_dataset/
  Contains the dataset curated for Experiment 1.

• vector_alignment_utils.py
  Implements the core alignment algorithms:

  • SPMC (Spherical Probabilistic Matching via Correlation)
  • FRS (Fast Rotation Search)
  • SPMC+FRS (Hybrid approach)

• vector_alignment_visualization_utils.py
  Includes utilities for visualizing the alignment process and evaluating results.

• exp1_example.ipynb
  Jupyter Notebook demonstrating:

  • Loading of source and template patterns
  • Selection of the desired algorithm
  • Execution of the registration pipeline with visual feedback

Installation

Prerequisites

• Python ≥ 3.8

Step 1: Clone the Repository

git clone https://github.com/<your-username>/<repository-name>.git
cd <repository-name>

Step 2: Install Dependencies

pip install -r requirements.txt